Spectroscopic sensor calibration system and method of calibration thereof

Abstract

The present application provides a spectral sensor calibration system and a calibration method thereof, wherein the calibration method comprises the following steps: S1, providing a light source assembly, wherein the light source assembly comprises a plurality of monochromatic light sources, and the brightness of at least one monochromatic light source is adjusted according to the acquisition frame rate and spectral response sensitivity of a spectral sensor to be calibrated; S2, controlling the monochromatic light sources to generate monochromatic light of different wave bands in turn, and ensuring that the spectral sensor to be calibrated acquires data at full frame rate; S3, receiving the monochromatic light by the spectral sensor to be calibrated and obtaining corresponding optical signal values; and S4, generating resource data based on the optical signal values, wherein the resource data is used for spectral recovery algorithm or sensor parameter calibration.

Description

Technical Field

[0001] This invention relates to the field of sensor technology, and in particular to a spectral sensor calibration system and calibration method thereof. Background Technology

[0002] In recent years, spectral detection technology has been increasingly widely used in consumer electronics, biomedicine, and industrial inspection. Computational spectral sensors, by combining optical modulation layer design with compressed sensing algorithms, have achieved breakthroughs in miniaturization, high resolution, and large bandwidth, gradually becoming a research hotspot in the industry. However, this technology faces a key bottleneck in its industrialization process: due to fluctuations in manufacturing processes and differences in material properties, mass-produced spectral sensors exhibit significant defects in spectral response consistency, specifically manifested as spectral shift and peak response deviation, directly affecting the accuracy of spectral reconstruction and the reliability of the equipment.

[0003] Traditional calibration systems employ a structure combining a beam splitter with a mechanical conveyor belt. This involves physically splitting the light to generate monochromatic light and then calibrating the sensor response point-by-point. This method has two inherent drawbacks: First, the beam splitter system is limited by the precision of mechanical movement, resulting in low calibration efficiency. Calibration of a single device can take several minutes, making it difficult to meet production line cycle time requirements. Second, the optical path stability is poor during beam splitting. Fluctuations in light source intensity and mechanical positioning errors lead to insufficient calibration repeatability. Actual measurement data shows that the correlation coefficient of the spectral response matrix varies significantly across multiple calibrations of the same model of equipment. Furthermore, traditional solutions have limited calibration capabilities for broadband composite light sources and cannot effectively correct the sensor's nonlinear response characteristics, resulting in spectral reconstruction errors that fail to meet requirements in practical applications.

[0004] While existing technologies have proposed specific optical modulation layer designs, they have not solved the calibration standardization problem in mass production. Therefore, there is an urgent need to develop a new calibration system with high-speed calibration capabilities and a high-precision compensation mechanism to overcome industrialization bottlenecks and ensure the measurement consistency of spectral sensors under complex operating conditions.

[0005] The description herein provides only background information in relation to the present invention and does not necessarily constitute prior art. Summary of the Invention

[0006] A key advantage of this invention is that it provides a spectral sensor calibration system and calibration method thereof, wherein the spectral sensor calibration system can acquire resource data band by band, improve the quality of resource data, and thereby reduce the difficulty and error of solving the problem.

[0007] Another advantage of the present invention is that it provides a spectral sensor calibration system and calibration method thereof, wherein the spectral sensor calibration system is based on a quantification model and compensation algorithm for manufacturing errors, automatically corrects spectral shifts and response deviations, and reduces calibration repeatability errors.

[0008] Another advantage of the present invention is that it provides a spectral sensor calibration system and calibration method, wherein the spectral sensor calibration system adopts a broadband composite light source and a multi-dimensional calibration mode, which solves the problem that traditional monochromatic calibration cannot cover the nonlinear characteristics of the sensor.

[0009] Another advantage of this invention is that it provides a spectral sensor calibration system and calibration method thereof. The spectral sensor calibration system reduces reliance on high-precision spectrometers through optical path multiplexing design and modular architecture, thereby reducing equipment costs. At the same time, it supports simultaneous calibration of multiple sensors, which helps to improve production capacity.

[0010] Another advantage of this invention is that it provides a spectral sensor calibration system and calibration method, wherein the spectral sensor calibration system establishes a wide-spectrum nonlinear calibration model and achieves full-range characteristic coverage through multi-order response fitting technology, thereby improving the reconstruction accuracy under complex lighting scenarios.

[0011] According to one aspect of the present invention, a spectral sensor calibration method of the present invention, which can achieve the foregoing and other objects and advantages, includes the following steps:

[0012] S1, providing a light source assembly, wherein the light source assembly includes multiple monochromatic light sources, and the brightness of at least one monochromatic light source is adjusted according to the acquisition frame rate and spectral response sensitivity of the spectral sensor to be calibrated;

[0013] S2, control the monochromatic light source to generate monochromatic light of different wavelengths in sequence, and ensure that the spectral sensor to be calibrated collects data at full frame rate;

[0014] S3, receiving the monochromatic light through the spectral sensor to be calibrated and acquiring the corresponding optical signal value; and

[0015] S4. Based on the optical signal value, resource data is generated, and the resource data is used for spectral recovery algorithms or sensor parameter calibration.

[0016] According to an embodiment of the present invention, in step S2, the rule for brightness adjustment is as follows: if the sensor cannot collect data at full frame rate within the minimum exposure time under the current brightness, the brightness is gradually increased according to a preset gradient until the full frame rate collection condition is met.

[0017] According to one embodiment of the present invention, step S3 further includes monitoring the spectral and intensity data of the monochromatic light source by at least one standard detector; the standard detector is a spectrometer, illuminance meter or luminance meter.

[0018] According to one embodiment of the present invention, the standard detector monitors in real time whether there is a deviation in the light generated by the light source assembly.

[0019] According to an embodiment of the present invention, in step S4, the resource data includes at least one of the following: spectral response curve, layer parameters for spectral recovery in a neural network model, and end-to-end consistent resource files generated based on a large model for improving spectral consistency across all scenarios.

[0020] According to one embodiment of the present invention, step S5 is further included:

[0021] The spectral data of the light source component are verified using a first detector and a second detector.

[0022] In this configuration, the first detector receives optical signals synchronously with the spectral sensor to be calibrated, and the second detector collects data in a point inspection mode.

[0023] By comparing the data from the sensor, the first detector, and the second detector, the system is identified as abnormal and a calibration or alarm is triggered.

[0024] According to another aspect of this application, this application further provides a spectral sensor calibration system, including...

[0025] System main body;

[0026] A light source assembly includes multiple monochromatic light sources, wherein at least two monochromatic light sources have different brightness levels; and

[0027] The processing module is used to generate resource data required for spectral recovery algorithms or calibration parameters based on the light signal values ​​received by the spectral sensor to be calibrated.

[0028] According to one embodiment of the present invention, a standard detector is further included, which is used to monitor, or as needed, whether there is a deviation in the light generated by the light source assembly.

[0029] According to one embodiment of the present invention, the standard detector is implemented as an illuminance meter, a luminance meter, or a spectrometer.

[0030] According to one embodiment of the present invention, the light source assembly includes at least eight monochromatic light sources.

[0031] According to one embodiment of the present invention, the full width at half maximum (FWHM) of each monochromatic light source is ≤50nm, and the peak wavelength spacing between any two monochromatic light sources is ≥1 / 5 of the sum of the FWHMs of the two monochromatic light sources.

[0032] According to one embodiment of the present invention, the wavelength spacing between the spectral peaks of any two monochromatic light sources is greater than or equal to half the sum of the full width at half maximum (FWHM) of the two monochromatic light sources.

[0033] According to one embodiment of the present invention, the light source component is a composite light source.

[0034] According to one embodiment of the present invention, the standard detector includes a first detector and a second detector, wherein the first detector synchronously receives the light signal of the light source component with the spectral sensor to be calibrated; the second detector periodically collects data in a point inspection mode and cross-verifies it with the data of the first detector and the sensor to determine system abnormalities.

[0035] According to one embodiment of the present invention, the first detector is a spectrometer, and the second detector is a spectrometer, a luminance meter, or an illuminance meter.

[0036] According to one embodiment of the present invention, the first detector is a spectrometer, and the second detector is a luminance meter or illuminance meter.

[0037] The further objects and advantages of the invention will become fully apparent from the following description and accompanying drawings.

[0038] These and other objects, features and advantages of the present invention will become fully apparent from the following detailed description and accompanying drawings. Attached Figure Description

[0039] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. In the drawings, unless otherwise specified, the same reference numerals are used to denote the same parts. Wherein:

[0040] Figure 1A and Figure 1B This is a schematic diagram of the structural framework of a spectral sensor calibration system according to a first preferred embodiment of the present invention.

[0041] Figure 2 This is a schematic diagram of the calibration method steps of the spectral sensor calibration system according to the first preferred embodiment of the present invention. Detailed Implementation

[0042] It should be noted that the embodiments shown in the accompanying drawings are merely examples used to specifically and vividly explain and illustrate the concept of the present invention. They are not necessarily drawn to scale in terms of size and structure, nor do they constitute a limitation on the concept of the present invention.

[0043] The directional terms such as up, down, left, right, front, back, front, back, top, and bottom mentioned or possibly used in this specification are defined relative to the structures shown in the various accompanying drawings. They are relative concepts and may therefore vary depending on their location and usage. Therefore, these or other directional terms should not be interpreted as restrictive.

[0044] It is understood that the term "a" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple, and the term "a" should not be understood as a limitation on the number.

[0045] The spectral sensor to be calibrated in this application includes a filter structure and an image sensor. The filter structure is located on the photosensitive path of the image sensor and is a broadband filter structure in the frequency domain or wavelength domain. The pass spectra of different wavelengths of the filter structure are not completely the same at different locations. The filter structure can be a metasurface, photonic crystal, nanopillar, multilayer film, filter material, quantum dot, MEMS (microelectromechanical systems), FP etalon, cavity layer, waveguide layer, diffraction element, or other structures or materials with filtering properties. For example, the filter structure can be the light modulation layer in Chinese patent CN201921223201.2, or it can be a specific filter material formed on the image sensor. The filter material can be a dye or pigment. The spectral sensor can include optical components, such as a lens group, a lens group and a filter, or a light homogenizer.

[0046] The image sensor can be a CMOS image sensor (CIS), CCD, array photodetector, etc. Additionally, the spectral sensor may include a data processing unit, which can be a MCU, CPU, GPU, FPGA, NPU, ASIC, etc., capable of exporting the data generated by the image sensor for external processing.

[0047] For example, after the image sensor measures the spectral response, it is fed into the data processing unit for reconstruction calculations. This process is described in detail below:

[0048] The intensity signals of incident light at different wavelengths λ are denoted as x(λ), and the transmission spectrum matrix of the filter structure is denoted as T(λ). The filter (filter structure) has m sets of structural units, and the transmission spectrum of each set of structural units is different. Overall, the filter structure can be denoted as Ti(λ) (i=1,2,3,…,m). Each set of structural units has a corresponding physical pixel below it, which detects the light intensity bi modulated by the filtered light structure. In a specific embodiment of this application, one physical pixel corresponds to one set of structural units, but it is not limited to this. In other embodiments, multiple physical pixels can also be grouped together to correspond to one set of structural units. At least two sets of structural units constitute a "spectral pixel" (which can be understood as multiple sets of structural units and corresponding image sensors constituting a spectral pixel).

[0049] It should be noted that the number of effective transmission spectra (the transmission spectrum used for spectral reconstruction is called the effective transmission spectrum) Ti(λ) of the filter structure may not be the same as the number of structural units. The transmission spectrum of the filter structure is manually set, tested, or calculated according to certain rules based on the needs of identification or reconstruction (for example, the transmission spectrum obtained by testing each structural unit is the effective transmission spectrum). Therefore, the number of effective transmission spectra of the filter structure may be less than or even more than the number of structural units. In this modified embodiment, a certain transmission spectrum curve is not necessarily determined by a group of structural units. Furthermore, the present invention can use at least one spectral pixel to reconstruct an image. That is, the spectral sensor mentioned in this application can reconstruct the spectral curve or perform spectral imaging based on the spectral response curve.

[0050] The relationship between the spectral distribution of incident light and the measurements from the image sensor can be expressed by the following formula:

[0051] bi=∫x(λ)*Ti(λ)*R(λ)dλ

[0052] After discretization, we get

[0053] bi=Σ(x(λ)*Ti(λ)*R(λ))

[0054] Where R(λ) is the response of the image sensor, denoted as:

[0055] Ai(λ)=Ti(λ)*R(λ),

[0056] The above equation can then be extended into matrix form:

[0057]

[0058] Where bi (i = 1, 2, 3, ..., m) is the response of the image sensor after the light transmission filter structure to be calibrated, corresponding to the light signal values ​​of the image sensor corresponding to each of the m structural units. When one physical pixel corresponds to one structural unit, it can be understood as the light signal values ​​corresponding to m physical pixels, which is a vector of length m. The spectral response curve A of the spectral sensor is the system's response to light of different wavelengths, determined by the transmittance of the filter structure and the quantum efficiency of the image sensor. The spectral response curve A is a matrix, where each row vector corresponds to a set of structural units responding to incident light of different wavelengths. Here, the incident light is sampled discretely and uniformly, with a total of n sampling points. The number of columns in A is the same as the number of sampling points for the incident light. Here, x(λ) is the light intensity of the incident light at different wavelengths λ, which is the incident light spectrum to be calibrated.

[0059] The spectral sensor calibration system in this application is used to acquire resource data of the spectral sensor during mass production. The resource data can be the spectral response curve of the spectral sensor, or other data used to recover the spectrum. It can be used to recover the spectrum, such as the parameters of one or more layers when using a neural network to recover the spectrum.

[0060] Refer to the accompanying drawings in this application specification. Figures 1A to 2 As shown, a spectral sensor calibration system and calibration method according to a first preferred embodiment of this application will be described below. The spectral sensor calibration system includes a system body 10, a light source assembly 20, and a detection station 30 disposed on the system body 10. The light source assembly 20 is fixed to the system body 10 and is disposed opposite to the detection station 30. The spectral sensor to be calibrated is placed on the detection station 30, so that the calibrated spectral sensor placed on the detection station 30 can receive the light emitted by the light source assembly 20.

[0061] The light source assembly 20 includes multiple monochromatic light sources 21, for example, 8 or more, or even 13, 16, 19, 23, 27, 30, etc. It is worth noting that the more monochromatic light sources 21 in the light source assembly 20, the more accurate the acquired resource data. Each monochromatic light source in the light source assembly 20 has a different center wavelength. Preferably, the full width at half maximum (FWHM) of the monochromatic light source 21 in the light source assembly 20 is less than or equal to 50 nm; wherein, the wavelength interval between the spectral peaks of any two monochromatic light sources 21 is greater than one-fifth of the sum of the FWHMs of the two monochromatic light sources 21. The monochromatic light sources 21 can acquire resource data band by band, which can improve the quality of resource data. For example, when the resource data is implemented as a spectral response curve, it can be considered as performing spectral response curve correction band by band, which can reduce the difficulty and error of the solution. Preferably, there are at least two monochromatic light sources 21 whose wavelength interval between the spectral peaks is greater than half of the sum of the FWHMs of the two monochromatic light sources 21. In certain embodiments of this application, the light source component 20 may be implemented as a composite light source, i.e., with a full width at half maximum (FWHM) greater than 50 nm, or with multiple peaks, or with a gradually changing peak.

[0062] It is worth mentioning that, based on the relative sensitivity of the spectral sensor to be calibrated when acquiring monochromatic light data at different wavelengths, it is necessary to design the brightness of monochromatic light sources in different wavelength bands so that the spectral sensor to be calibrated can acquire data at full frame rate with a sufficiently short exposure time under the light source assembly, thereby improving efficiency and accuracy. That is, at least two of the monochromatic light sources 21 generate light with different brightness, and the brightness is determined by the acquisition frame rate and spectral response sensitivity of the spectral sensor to be calibrated. When the brightness is insufficient to ensure that the spectral sensor to be calibrated acquires data at full frame rate, the brightness is appropriately increased. In this preferred embodiment of the present application, the brightness of at least one monochromatic light source 21 of the light source assembly 20 is adjustable.

[0063] In this preferred embodiment of the present application, the monochromatic light source 21 can be implemented as a light source such as an LED or a laser. Furthermore, the monochromatic light source 21 can acquire resource data band by band to improve the quality of the resource data. For example, when the resource data is implemented as a spectral response curve, band-by-band spectral response curve correction can reduce the difficulty and error of the solution. Preferably, the monochromatic light source 21 of the present application is an orthogonal light source, that is, the wavelength interval between the spectral peaks of any two monochromatic light sources 21 is greater than or equal to one-fifth of the sum of the full width at half maximum (FWHM) of the two monochromatic spectra. Preferably, the wavelength interval between the spectral peaks of any two monochromatic light sources 21 is greater than or equal to one-half of the sum of the full WHM of the two monochromatic light sources 21.

[0064] The spectral sensor calibration system further includes at least one standard detector 40, wherein the standard detector 40 is fixed to the system body 10 to receive the light generated by the light source component 20, thereby enabling monitoring of the operation of the light source component 20.

[0065] The spectral sensor calibration system further includes a processing module 100, wherein the processing module 100 is connected to the spectral sensor and the light source assembly 20 to be calibrated, wherein the processing module 100 is used to control each of the monochromatic light sources of the light source assembly 20 to emit specific light, and to generate resource data required for spectral recovery algorithms or calibration parameters based on the light signal values ​​of the spectral sensor.

[0066] The light generated by the monochromatic light source 21 of the light source assembly 20 is uniform light reaching the spectral sensor. Without considering any light-blocking effect, this ensures that the entire target surface of the spectral sensor receives light of the same intensity and spectrum, thus enabling uniform calibration across the entire target surface. The spectral sensor calibration system further includes a light-diffusing component 50, which is positioned at the front end of the light source assembly 20. That is, the light emitted by the light source assembly 20 passes through the light-diffusing component 50 before reaching the spectral sensor to be calibrated. It is worth noting that the light-diffusing component 50 can be, but is not limited to, a microlens array, a light-diffusing sheet, etc.

[0067] The spectral sensor calibration system further includes at least one detection station 30, which is arranged opposite to the light source assembly 20. The spectral sensor to be calibrated is set at the detection station 30, so that the light generated by the light source assembly 20 when it is working can be received by the spectral sensor to be calibrated.

[0068] Furthermore, the different monochromatic light sources 21 of the light source component 20 are generated according to requirements. For example, they can generate corresponding monochromatic light according to the center band from small to large, or they can generate it from large to small, or they can generate it according to requirements.

[0069] It is worth mentioning that, in this preferred embodiment of the application, the standard detector 40 is used to monitor, or as needed, whether there is a deviation in the light generated by the light source component. Preferably, it can be monitored in real time or periodically, or it can be monitored intermittently as needed. The standard detector 40 can be implemented as an illuminance meter, a luminance meter, or a spectrometer, etc., a detection device capable of detecting spectrum and intensity. The spectral sensor calibration system further includes at least one placement fixture 70, wherein the placement fixture 70 can simultaneously fix at least one of the spectral sensors to be calibrated, that is, the placement fixture 70 is fixed to the detection station 30 and is used to fix the spectral sensor to be calibrated. In the embodiments of the application, preferably, the standard detector 40 can also be disposed in the system body 10 of the calibration system.

[0070] The standard detector 40 further includes a first detector 41 and a second detector 42, wherein the first detector 41 and the second detector 42 monitor the light source assembly to ensure that the calibration system is in normal working order when turned on and / or during operation. Specifically, the accurate identification of abnormalities in the light source, detector, and spectral sensor to be calibrated is ensured through mutual verification of the data from the first detector 41, the second detector 42, and the spectral sensor to be calibrated.

[0071] Specifically, the first detector 41 and the second detector 42 are used to detect the spectral data of the light generated by the light source component 20. At least one detector (such as the first detector 41) and the spectral sensor to be calibrated simultaneously receive the spectral data of the light source component, and then analyze the two spectral data to determine whether there is an anomaly. The other detector (such as the second detector 42) can be used for routine inspection, for example, when the light source component 20 is turned on, or randomly or according to regulations during operation, to receive the spectral data of the light source component 20 and compare it with the other two data to determine whether there is an anomaly. The first detector 41 and the second detector 42 can be, but are not limited to, a spectrometer, a luminance meter, or an illuminance meter. Preferably, in a specific example of this application, at least one of the first detector 41 or the second detector 42 is a standard detector, such as a spectrometer.

[0072] Refer to the accompanying drawings in this application specification. Figure 2 As shown, a spectral sensor calibration method according to another aspect of this application is illustrated in the following description, wherein the spectral sensor calibration method includes the following steps:

[0073] S1, a light source assembly 20 is provided, wherein the light source assembly 20 includes a plurality of monochromatic light sources 21, and the brightness of at least one monochromatic light source 21 is adjusted according to the acquisition frame rate and spectral response sensitivity of the spectral sensor to be calibrated.

[0074] S2, control the monochromatic light source 21 to generate monochromatic light of different wavelengths in sequence, and ensure that the spectral sensor to be calibrated collects data, preferably at full frame rate;

[0075] S3, receiving the monochromatic light through the spectral sensor to be calibrated and acquiring the corresponding optical signal value; and

[0076] S4. Based on the optical signal value, resource data is generated, and the resource data is used for spectral recovery algorithms or sensor parameter calibration.

[0077] It should be noted that in step S1 of the spectral sensor calibration method of this preferred embodiment, the light source assembly 20 consists of at least N monochromatic light sources, where N≥8. The monochromatic light source 21 is not an absolutely single wavelength, but rather refers to light with a full width at half maximum (FWHM) of less than or equal to 50 nm. For example, in some high-precision spectral detection applications, selecting a monochromatic light source with a smaller FWHM (such as 20 nm or 30 nm) can improve the resolution and accuracy of the detection. Simultaneously, the light source assembly 20 can also incorporate composite light sources according to actual needs. Composite light sources are characterized by a FWHM greater than 50 nm, or having multiple peaks (such as multiple emission regions containing different wavelengths, forming multiple peaks), or exhibiting a gradual change in light intensity. Such composite light sources play an important role in simulating complex natural light sources or specific industrial light source scenarios.

[0078] Monochromatic light sources can be implemented using common light sources such as LEDs (light-emitting diodes) or lasers. LEDs are widely used in the calibration of many spectral sensors due to their low cost, long lifespan, and ease of control; while lasers offer higher brightness and narrower spectral width, making them suitable for applications requiring extremely high spectral accuracy. Furthermore, monochromatic light sources can acquire resource data band by band, a characteristic crucial for improving the quality of this data. Taking the spectral response curve as an example, band-by-band acquisition means that each band can be calibrated more precisely, effectively reducing the difficulty and error of the solution compared to overall calibration.

[0079] Preferably, the monochromatic light source 21 is an orthogonal light source to improve the accuracy and effectiveness of calibration. Specifically, the orthogonality requirement is that the wavelength interval between the spectral peaks of any two monochromatic light sources is greater than or equal to one-fifth of the sum of their spectral full width at half maximum (FWHM). More preferably, the wavelength interval between the spectral peaks of any two monochromatic light sources is greater than or equal to one-half of the sum of their spectral FWHM. This configuration avoids spectral overlap interference between different monochromatic light sources, enabling the spectral sensor to more clearly distinguish light of different wavelengths, thereby improving the reliability of the calibration results. It should be noted that the orthogonality of the light sources in this application refers to the monochromatic light sources used in the calibration. Light sources that are present in the system but not actually used in the calibration process are not considered for orthogonality.

[0080] Furthermore, to ensure that the entire target surface of the spectral sensor receives light of uniform intensity and spectrum, thus achieving unified calibration across the entire target surface, the light generated by the monochromatic light source in the light source assembly should be uniform light reaching the spectral sensor. There are two ways to achieve uniform light: one is to optimize the design of the monochromatic light source itself so that the emitted light is uniform; the other is to place the aforementioned light-homing component between the light source assembly and the spectral sensor to be calibrated. Common light-homing components include microlens arrays, which can uniformly distribute incident light to different areas, making the light intensity reaching the spectral sensor more uniform. Light-homing sheets can also play a similar role, using principles such as scattering and refraction to uniformly distribute light within a certain range.

[0081] In step S2 of the spectral sensor calibration method of this preferred embodiment, different monochromatic lights are sequentially generated by the monochromatic light source of the light source assembly according to requirements. The brightness of the monochromatic light source needs to be designed for different wavelengths based on the relative sensitivity of the spectral sensor to be calibrated when acquiring data with monochromatic light at different wavelengths. This is because different spectral sensors have different response capabilities to light of different wavelengths; some wavelengths may be more sensitive to the sensor, while others may be relatively less sensitive. To ensure that the spectral sensor to be calibrated can acquire data at full frame rate with sufficiently short exposure times under the light source assembly, thereby improving efficiency and accuracy, the brightness of the light source needs to be adjusted. For example, if the sensor has a weak response to a certain wavelength of light, the brightness of the corresponding monochromatic light source needs to be appropriately increased to ensure that the sensor can receive sufficient light signals, thereby achieving full frame rate acquisition.

[0082] It is worth mentioning that in step S2 of the spectral sensor calibration method of this preferred embodiment, at least two monochromatic light sources generate light with different brightness, and the brightness is determined by the acquisition frame rate and spectral response sensitivity of the spectral sensor to be calibrated. When the brightness is insufficient to ensure that the spectral sensor to be calibrated acquires data at full frame rate, the brightness should be appropriately increased. In the actual calibration process, the brightness of the light source can be dynamically adjusted by monitoring the acquisition frame rate and light signal intensity of the spectral sensor to achieve the best calibration effect.

[0083] In step S3 of the spectral sensor calibration method of this preferred embodiment, the spectral sensor to be calibrated receives light generated by different monochromatic light sources and obtains corresponding light signal values. The spectral sensor to be calibrated is positioned relative to the light source assembly, and is located at a detection station so that the light generated by the light source assembly during operation can be received by the spectral sensor to be calibrated. This relative positioning ensures that light accurately illuminates the sensor, avoiding light deflection and loss.

[0084] Furthermore, the different monochromatic light sources of the light source component are generated according to requirements. For example, they can be generated according to the center band from small to large, so as to systematically cover the entire spectral range and facilitate subsequent analysis and processing of spectral data; or they can be generated from large to small center bands, which may be more conducive to data acquisition and analysis in certain specific application scenarios; or they can be generated according to requirements, such as prioritizing the generation of monochromatic light of certain wavelengths based on the degree of attention paid to certain specific wavelengths in actual applications.

[0085] The standard measuring instrument is used to monitor, or as needed, whether there are deviations in the light generated by the light source components. Preferably, real-time monitoring is possible. The standard measuring instrument can be an illuminance meter, a luminance meter, or a spectrometer, etc., that can detect the spectrum and intensity. For example, in high-precision spectral calibration, using a spectrometer as the standard measuring instrument can accurately detect the spectral characteristics of the light source and promptly identify problems such as wavelength drift or changes in spectral width.

[0086] In step S4 of the spectral sensor calibration method of this preferred embodiment, resource data is acquired through optical signal values. This resource data includes spectral recovery algorithms, software, or information related to the spectrum, algorithmic calculations of indicators, and resource files required by the software. The resource data may be a spectral response curve, which reflects the response characteristics of the spectral sensor to light of different wavelengths and is an important basis for evaluating the performance of the spectral sensor. The resource data may also be parameters required for the calculation of spectral recovery algorithms or information related to the spectrum. For example, when using neural networks to recover the spectrum, the parameters of one or more layers play a crucial role in accurately recovering the spectrum.

[0087] It is worth mentioning that there are multiple ways to obtain resource data from optical signal values. The calculation process can be performed directly on the calibration device, using the device's built-in processor and algorithms to process and analyze the optical signal values ​​and quickly obtain the required resource data; alternatively, it can be uploaded to the cloud for calculation, utilizing the cloud's powerful computing capabilities and storage resources to perform more complex and precise processing on large amounts of optical signal value data, while also enabling data sharing and remote monitoring.

[0088] Taking the consistency calibration model as an example, the raw images acquired by the spectral sensor, the acquired light source spectrum, and the pre-prepared data are used as training inputs. Then, a large-scale model for improving spectral consistency across all scenarios (based on a pre-trained large-scale model) is used for training. This large-scale model is an end-to-end model that can directly generate the required consistency resource files based on the data acquired by the spectral sensor. Simultaneously, the training results also include customizable evaluation metrics for certain scenarios to assess the consistency improvement. This process is highly robust to spectral deviations of the acquired light source, reducing the requirements for spectral consistency in the input data. Furthermore, it is highly customizable, allowing for regional variations, evaluation scenarios, and additional enhancements for specific scenarios. For example, in industrial production, different batches of light sources may have certain spectral differences. This large-scale model can effectively overcome the impact of these differences on spectral sensor calibration, improving the accuracy and consistency of the calibration.

[0089] In step S5 of the spectral sensor calibration method of this preferred embodiment, the spectral sensor to be calibrated is checked using at least two detectors (i.e., a first detector and a second detector). This check ensures that the calibration system is functioning normally during startup and / or operation. Specifically, mutual verification of the calibration data from the first detector, the second detector, and the spectral sensor to be calibrated ensures accurate identification of any abnormalities in the light source, spectrometer, or spectral sensor. Preferably, the first and second detectors are used to detect the spectral data of the light generated by the light source assembly, and at least one standard detector and the spectral sensor to be calibrated simultaneously receive the spectral data from the light source assembly, then the two sets of spectral data are analyzed to determine if any abnormalities exist.

[0090] For example, if the first detector and the spectral sensor to be calibrated simultaneously receive spectral data from the light source, and there is a significant difference between the two data, then there may be an abnormality in the light source, a sensor malfunction, or other problems. The second detector can then be used for routine checks, such as receiving the spectral data of the light source component randomly or according to regulations during operation, comparing and analyzing it with the other two data sets to determine if there are any abnormalities.

[0091] The first and second detectors can be standard detectors such as spectrometers, luminance meters, or illuminance meters. Preferably, one of the first or second detectors is one of these standard detectors. Different types of detectors have different functions and characteristics. Spectrometers can provide detailed spectral information, while luminance meters and illuminance meters are mainly used to measure light intensity. Based on actual calibration requirements and accuracy requirements, a suitable combination of detectors can be selected to ensure accurate detection of anomalies in the system.

[0092] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are merely examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functions and structural principles of the present invention have been demonstrated and explained in the embodiments, and any variations or modifications may be made to the implementation of the present invention without departing from the stated principles.

Claims

1. A method for calibrating a spectral sensor, characterized in that, The spectral sensor calibration method includes the following steps: S1, providing a light source assembly, wherein the light source assembly includes multiple monochromatic light sources, and the brightness of at least one monochromatic light source is adjusted according to the acquisition frame rate and spectral response sensitivity of the spectral sensor to be calibrated; S2, control the monochromatic light source to generate monochromatic light of different wavelengths in sequence, and ensure that the spectral sensor to be calibrated collects data at full frame rate; S3, receive the monochromatic light through the spectral sensor to be calibrated and obtain the corresponding optical signal value; as well as S4. Based on the optical signal value, resource data is generated, and the resource data is used for spectral recovery algorithms or sensor parameter calibration.

2. The method according to claim 1, characterized in that, In step S2, the brightness adjustment rule is as follows: if the sensor cannot collect data at full frame rate within the minimum exposure time under the current brightness, the brightness is gradually increased according to the preset gradient until the full frame rate collection condition is met.

3. The method according to claim 2, characterized in that, Step S3 further includes monitoring the spectral and intensity data of the monochromatic light source using at least one standard detector; the standard detector is a spectrometer, illuminance meter, or luminance meter.

4. The method according to claim 3, characterized in that, The standard testing instrument monitors in real time whether there are any deviations in the light generated by the light source component.

5. The method according to claim 3, characterized in that, In step S4, the resource data includes at least one of the following: spectral response curves, layer parameters for spectral recovery in a neural network model, and end-to-end consistent resource files generated based on a large model for improving spectral consistency across all scenarios.

6. The method according to claim 1, characterized in that, It also includes step S5: The spectral data of the light source component are verified using a first detector and a second detector. In this configuration, the first detector receives optical signals synchronously with the spectral sensor to be calibrated, and the second detector collects data in a point inspection mode. By comparing the data from the spectral sensor, the first detector, and the second detector, the system is identified as abnormal and a calibration or alarm is triggered.

7. A spectral sensor calibration system, characterized in that, include System main body; A light source assembly includes multiple monochromatic light sources, wherein at least two monochromatic light sources have different brightness levels; as well as The processing module generates resource data required for the spectral recovery algorithm or calibration parameters based on the light signal values ​​received by the spectral sensor to be calibrated.

8. The system according to claim 7, characterized in that, It further includes a standard testing instrument used to monitor, or as needed, whether there are deviations in the light generated by the light source components.

9. The system according to claim 8, characterized in that, The standard testing instrument is implemented as an illuminance meter, luminance meter, or spectrometer.

10. The system according to claim 7, characterized in that, The light source assembly includes at least eight monochromatic light sources.

11. The system according to claim 10, characterized in that, The full width at half maximum (FWHM) of each monochromatic light source is ≤50nm, and the peak wavelength spacing between any two monochromatic light sources is ≥1 / 5 of the sum of the FWHMs of the two monochromatic light sources.

12. The system according to claim 10, characterized in that, The wavelength interval between the spectral peaks of any two of the monochromatic light sources is greater than or equal to half the sum of the full width at half maximum (FWHM) of the two monochromatic light sources.

13. The system according to claim 7, characterized in that, The light source component is a composite light source.

14. The system according to claim 10, characterized in that, The standard testing instrument includes a first testing instrument and a second testing instrument, wherein the first testing instrument synchronously receives the light signal of the light source component with the spectral sensor to be calibrated; the second testing instrument periodically collects data in a point inspection mode and cross-verifies it with the data of the first testing instrument and the sensor to determine system abnormalities.

15. The system according to claim 14, characterized in that, The first detector is a spectrometer, and the second detector is a spectrometer, luminance meter, or illuminance meter.

Images